Romanization of Bulgarian

An older system in the tradition of common Slavic scientific transliteration was adopted by the Council of Orthography and Transcription of Geographical Names in Sofia in 1972 and subsequently by the UN in 1977.[4] It is identical to that codified in the ISO norm ISO/R 9:1968. This system uses diacritic letters (⟨č, š, ž⟩) as well as ⟨j⟩ and ⟨c⟩. It was adopted in 1973 as the Bulgarian state standard BDS 1596:1973 which, although still valid formally[5] is no longer used in practice,[6] having been superseded by the 2009 Transliteration Act.

The second system was a French-oriented transliteration of personal and place names in the documents issued by the Bulgarian Ministry of Interior for travel abroad, used until 1999.[7]

Systems based on a radically different principle, which avoids diacritics and is optimized for compatibility with English sound-letter correspondences,[8] have come into official use in Bulgaria since the mid-1990s. These systems characteristically use ⟨ch, sh, zh⟩ rather than ⟨č, š, ž⟩, and ⟨y⟩ rather than ⟨j⟩.

One such system was proposed in Danchev et al.'s English Dictionary of Bulgarian Names of 1989.[9]

A similar system (differing from the former in the treatment of letters ъ, у, and digraphs ай, ей, ой and уй), called the "Streamlined System" by Ivanov (2003)[8] and Gaidarska (1998),[10] was adopted in 1995[11] for use in Bulgarian-related place names in Antarctica by the Antarctic Place-names Commission of Bulgaria.[12] Another system along similar lines, differing from the Antarctic one only in the treatment of ц (⟨ts⟩ vs. ⟨c⟩), was adopted by the Bulgarian authorities for use in identity documents in 1999;[8] after an amendment in 2000, the official Bulgarian system became identical with that of the Antarctica Commission.

The new official Bulgarian system does not allow for unambiguous mapping back into Cyrillic, since unlike most other systems it does not distinguish between ъ and а (both rendered as a). It also does not distinguish between the digraph values of ⟨zh=ж⟩, ⟨sh=ш⟩ and the value of the same Roman strings in rendering accidental clusters of separate Cyrillic letters ⟨zh=зх⟩ and ⟨sh=сх⟩, as they occur in words like изход (izhod) or схема (shema).[6]

A modification of the system using a diacritic was proposed in the authoritative New Orthographic Dictionary of the Bulgarian Language[13] in 2002, with ъ rendered as ă rather than a. However, that proposal was not adopted for official usage, and failed to become established in popular practice.

An exception to the rules was introduced by the Bulgarian authorities in 2006, mandating the transliteration of word-final -ия as -ia rather than -iya in given names and geographical names (such as Ilia, Maria and Bulgaria, Sofia, Trakia etc.).[14][15][16] In 2009, a law passed by the Bulgarian parliament made this system mandatory for all official use and some types of private publications, expanding also the application of the ia-exception rule to all -ия in word-final position.[1]

The Streamlined system was adopted by UN[17][18] in 2012, and by BGN and PCGN[19][20] in 2013.

According to Arenstein, "The international roots of the Bulgarian romanization system strike at the core of one of romanization's most perplexing paradoxes: an impulse to redefine and distinguish national identity while also ensuring the accessibility of that identity to outside groups. In other words, instilling nationalism with a sense of internationalism."[3]

A variant of the Streamlined System allowing for unambiguous mapping back into Cyrillic was proposed by Ivanov, Skordev and Dobrev in 2010 to be used in cases when the retrieval of the original Cyrillic forms is essential.[6] For that purpose, certain Cyrillic letters and combinations of letters are transliterated as follows: ъ→`a, ь→`y, зх→z|h, йа→y|a, йу→y|u, сх→s|h, тс→t|s, тш→t|sh, тщ→t|sht, шт→sh|t, шц→sh|ts, ия (in final position, if the ia-exception rule is applied) →i|a. The standard transliteration form of a given text is obtained from its unambiguously reversible one by simply removing the additional symbols ` and |.

Systems along similar lines to the new official Bulgarian system, though with differences regarding the letters х, ъ, ь, ю and я, have also been in use in the ALA-LC Romanization scheme of the Library of Congress, British Standard 2959:1958, the now-superseded 1952 BGN/PCGN romanization of the United States and British geographic naming institutions, and the 1917 system of the British Academy.[21]

The ISO 9 standard, in its 1995 version, has introduced another romanization system that works with a consistent one-to-one reversible mapping, resorting to rare diacritic combinations such as ⟨â,û,ŝ⟩.

The GOST 7.79-2000 "Rules of transliteration of Cyrillic script by Latin alphabet" contains an unambiguous and reversible ASCII-compatible transliteration system for Bulgarian: й→j, х→x, ц→c or cz, ч→ch, ш→sh, щ→sth, ъ→a`, ь→`, ю→yu, я→ya.